Human alpha-1-proteinase inhibitor (A1-PI), which is a serine protease inhibitor and which in vivo inhibits the activity of neutrophil elastase, undergoes guanidine-HCl (Gu) induced biphasic unfolding with an intermediate state in 1.4 M Gu. Previously, we demonstrated that in 1.4 M Gu over 1 h at 23 deg C an apparent equilibrium forms among a monomeric and polymeric intermediates. More recently, with highly purified A1-PI, Dr. Marszal has shown that the apparent equilibrium involves a monomeric and a dimeric intermediate only and that the folded dimer slowly polymerizes to form high molecular weight polymers, which appear to be multimers of dimer. In the presence of folded monomer, this dimer polymerizes with no apparent change in concentration of monomer thereby indicating that dimer is, in fact, is the polymerizing unit. A long proposed model of polymerization hypothesizes the insertion of the flexible, reactive site loop of A1-PI into a beta sheet of a second molecule and so on, i.e. loop-sheet (or head-to-tail) polymerization of monomer, to form linear polymers. To explain our observations, we now propose that the dimer is a symmetrical, head-to-head dimer, possibly of a reciprocal loop-sheet nature, that polymerizes to form linear polymers by tail-to-tail contact presumably of identical hydrophobic patches (possibly beta sheet structures) on the surface of the dimer. The loop-sheet model of polymerization was proposed partly on the basis of studies of thermally induced polymerization. Dr. Marszal has preliminary, limited proteolysis data on Gu-induced and heat-induced dimers that indicate different structures for these two dimers thereby suggesting that the mechanisms of formation may be different also. Since starting as a Visiting Associate in August 2000, Dr. Du has begun a program to establish the potency and stability of the CBER reference standard for A1-PI, which was manufactured under contract in 1988 but which has never been utilized since the potency is somewhat lower than the labeled value.